The Agent Skills Directory

[SAFE]: The skill provides structured guidelines for delegating tasks to specific sub-agents (investigator, builder, and reviewer). These sub-agents are designed to return formatted, high-density results to save on context window usage.
[SAFE]: No suspicious commands, network operations, or sensitive file access patterns were found. The skill focuses on workflow management and output formatting protocols.
[SAFE]: The skill includes a 'Automatic Release' clause where sub-agents are instructed to revert to standard language for security warnings or high-risk operations, showing a consideration for safety constraints.

genshijin-crew